May I ask two questions:-
1. Why are we discussing spins when no one knows whether the Air Asia aircraft ever got into a spin?
2. How many large commercial jet aircraft have ever got into a spin?
It seems to me that the discussion is taking a wrong turn.

The closest analogue might be a Russian airliner (Tu-154M) that got caught in a massive thunderstorm over Ukraine in 2006.

After large pitch oscillations due to the severe turbulence, the aircraft climbed beyond its service ceiling before entering a flat spin -- plunging 37,000ft into the ground. All 170 aboard perished.

Deep Stalls
 

I was always under the impression that entering a deep stall, where elevator authority is lost because airflow over the HS is in a "shadow" created by the wings, occurred mostly in high tail designs. Am I wrong?

Since the discussion on stall/spin is continuing, it would be useful to know if there is an answer to the following question. For the A320 is it known if there have been any tests, including wind tunnel tests during the design phase, that might indicate if the aircraft can be put into a deep stall condition, which in this case would be due to the THS itself stalled so that there is no longer elevator control effectiveness (since with a non-T-tail the mainplane would be unlikely to be able to shield the THS from normal airflow). In this case would this likely be unrecoverable even with significant height loss, or might there be unusual recovery techniques such as using thrust to provide airflow over the THS or possibly attempt to bank using residual aileron action, and then apply rudder to push the nose down enough to regain airflow across the mainplane?

This will likely stretch the limits of knowledge in an area beyond the normal flight envelope but it would be interesting to know if there is any knowledge of this extreme flight condition.

There is video of stall tests in flight for the 737 such as at https://www.youtube.com/watch?v=a4V8W31YwTQ, and presumably analogous tests for the regime as the stall begins were also conducted by Airbus for each of the type it designed and built? There is also a nice lecture given jointly by Boeing and Airbus about stall testing at:

https://www.youtube.com/watch?v=HVt6LiDbLos

1 The answer to #1 is that it appears to fit the impact conditions very well. See http://www.pprune.org/rumours-news/553569-air-asia-indonesia-lost-contact-surabaya-singapore-109.html#post8830502 At this point, it is just a theory until confirmed or denied by the accident investigation.
2. If it has wings, it can depart into a spin. Ask yourself what would happen to your twin underwing engine jet if it pitched up for some inappropriate reason and then compressor stalled one engine at an inopportune time due to the high transient AOA.

I know what will happen to at least 95% of posts on this thread when FDR readout is available.

I'm not sure what your point is, but all the Airbus prototypes I've been on have had crew escape hatches. They would be used in case something went very badly during any of the many stall tests that were done.

Thcrozier, while this definition is not universal, this is the way I think of stalls:
Stall - When the airflow separates causing the wing to loose lift. Most stalls can be readily recovered with a simple nose-down command, although the details can be tricky since one wing will often stall first resulting a sudden roll.
Deep stall - extreme stall where the nose is not allowed to drop and most forward speed is lost - forward speed is similar to the descent speed (e.g. Air France A330). Deep stalls take a long time to recover from since there is initially little airflow over the tail to provide a nose down command.
Unrecoverable stall - stall where the spoiled flow from the stalled wing effectively blanks the tail, completely preventing the necessary nose down command to recover - quite common in 'T' tail aircraft.


OK465 - I don't know how the FAA pilot would have done that, but my suspicion is that he was trying something that he wasn't supposed to and/or hadn't been briefed (that's happened before, where I do have first hand knowledge, when the FAA pilot performed a maneuver he'd been specifically told not to perform - tends to result in rather strained relations between Boeing and the Feds - and would also be consistent with it being kept 'quiet').

Why would one think that the investigators would not analyse previous flight data? For the purpose of a full investigation this must be done... simple as that. They're not "expecting" anything per se...

all this talk of stalls...

just remember for those of you who haven't heard of the many types of stall...if you can't get the nose down...

you might try rolling into a steep bank and this may cause the nose to come down eventually. this was taught many years ago when concepts like deep stall came out.


I would like to think that most planes certified will likely NOT get into the hairiest of stalls without some coaxing from either a very good test pilot or an odd sort of computer scenario.

I imagine , right now, that at least a dozen people in indonesia have a really good idea of what happened ...and what didn't happen...we might hear something in 10 days or so...just a guess

@A0283

Having had the misfortune to have done some transcripts of fatal air accidents, I can assure you that it takes longer than you would think and is more draining than you would think - especially when you know the people involved. Add in chaotic ECAM and cavalry charges and alarms etc etc and I am surprised that they have done as much as they have.
All the time you are aware that someone will challenge your transcripts so you go over and over. Each alarm has to be identified, each sound identified. Not a pleasant task at all.

in a number of these accidents we hear of the mayhem of alarms going off on the flight deck. no doubt making it even more difficult for pilots to try to think clearly. while audible alerts are obviously helpful if one goes off, they rapidly achieve the opposite of the desired effect in a full on crisis.

just thinking out loud, but does it makes sense at some point to have them shut off when they system senses the pilots are trying to respond? they obviously know they need to recover but being screamed at by hal non stop is not delivering any kind of useful infromation at that point.

We have reason to believe that the stall alarm cutting in and out on AF447 may have led to it being disregarded as spurious (and may have induced the PF to pull back again when airspeeds became valid and the alarm restarted). Certainly there is much more work to be done in providing prioritising relevant information to pilots in an audio saturated environment. I am anticipating a near impossible workload in the QZ8501 cockpit.

That's a new one

A recent sim exercise had an alarm repeat every couple of minutes during a busy procedure which rightly required the crews to formally acknowledge and evaluate the damned alarm each time it triggered. You don't want to get into the habit of blithely waving off repeat alarms without at least assuring yourself that it isn't something new, or worse, something new dressed in the previously used alert.

Cognitive overload under stress is a huge problem which the designers address by adding more unthinking and contextually ignorant alarms. Kind of like fixing a language problem by shouting.

Boeing and Airbus on Stall Testing
 

A March 2013 lecture by an test piloe and a flight test engineer from Boeing and Airbus on stall testing:

Royal Aeronautical Society | Event | Jet Transport Stalls

Royal Aeronautical Society | Podcast | Jet Transport Stalls

Video of the lecture, about an hour and 40 minutes long:

https://www.youtube.com/watch?v=HVt6LiDbLos

The individual speakers (times approximate):

Paul Bolds-Moorehead, Boeing flight test engineer, 0:00:00 to 0:02:30; 0:27:00 to 0:37:00; 1:31:00 to 1:37:00

Stephane Vaux, Airbus flight test engineer, 0:02:30 to 0:27:00; 0:37:00 to 0:49:00

Van Chaney, Boeing test pilot, 0:49:00 to 0:59:00

Terry Lutz, Airbus Test Pilot, 0:59:00 to 1:31:00

A similar set of briefing slides from 6 months later at a Society of Flight Test Engineers meeting:

http://sfte2013.com/files/78988645.pdf

A search engine turned up this description of the briefing, with a few web links that have since changed:

http://theairlinewebsite.com/topic/3...nsport-stalls/

The same Airbus and Boeing people published a journal paper later, but it's subscription only:

Royal Aeronautical Society | Aeronautical Journal | Stalling transport aircraft

Bootleg source:

http://www.sfte2013.com/files/75234188.pdf

Alarm
 

In modern Electrical Power stations, in case of serious problem for the security of the plant, the system sounds the alarm requiring immediate action from the operator. Afterwards, lesser urgent requests for action appears on the screen and so on...

It had been found that too much alarm sounding at the same time stress the operator which might not take the corrective actions as requested.

@ZFT


'audio saturated environment'
That's a new one

Not at all really and a lot of people, males in particular, have problems assimilating several noise sources.
It is something which happens at puberty to about 30 percent of males.
For myself it made learning morse code (in the Grey Funnel Line) the devils own job; all I heard was a stream of noise.
The same may happen if several people are talking at the same time; perhaps in somewhere like a pub where music may also be playing.
I wonder if this surprisingly common condition has been sufficiently thought about.

Yes it is quite clear to see that! I have seen guys & girls cancel genuine warnings because they are so used to cancelling a nuisance warning at a particular phase of flight....

All the talk on this thread of stalling may or may not be related to the accident in question, we will soon find out, but let me give the non pilots some practical info.

Is it possible to get a stall warning and a high speed warning go off at once without it being spurious?
Yes! Coffins corner.
All it would take is an updraft, especially when in bad weather and in turbulance. Descending to a lower altitude with increased margins between low speed and high speed is the only thing that will save your neck.

Put yourself in this situation. In moderate / severe turbulence at high level naturally only a few knots from low speed stall due to the G. Static tubes /smart probes ice up due to super cooled water droplets, first indication is that you are over speeding. If either you or the automatics reduce the power you are in a whole world of trouble.
To save your neck, recognition of the situation must be instant which is most likely not a practical stance.

You realise within a few seconds 'oh we must have unreliable airspeed' you decide to fly a sensible attitude and power as per the book, but guess what? it's already too late as you've already stalled the wing due to the initial reduction of power and / or severe turbulence.
So you now need to initiate a stall recovery but by this stage you don't trust any of your instruments so continue to fly that sensible attitude and power which you have been trained to do. It's very likely the instruments will be telling you that you are flying straight and level due to frozen ports. You are actually stalled and descending at 7000 + ft/min and you or the plane can't make sense of all the conflicting information before your eyes and that unfortunately is game over.

A good crew will recover from the above situation but the best solution is avoidance. Avoid the weather ahead, don't attempt to climb over and if you must go through or if you hit nasty turbulence, descend to a lower altitude immediately where your margins are increased and you buy yourself time to recognise the problem and carry out the sensible actions before it's already too late.

This is why a human brain will always be required in a flight deck to analyse and pre empt future problems.
Pilotless airliners? No thanks :ok:

Also worth a reminder that when humans become "maxed out" the brain "deletes" hearing.

CVR transcription is a very difficult task. There are typically four channels: LHS, RHS, Intercom & Cockpit ambient. Within these channels are multiple inputs such as Com 1, Com 2, Nav., Ramp, Inter-phone, PA, warnings etc. The comm. and nav. channels are easy to decode. The difficult ones are voice, ambient noises and warnings. These will have to be cross-checked against each other, FDR data, software versions, aircraft manuals etc. to determine exactly what warning sounded when, who heard and it and what was said. Only when combined with SOPs, cultural norms, company standards etc. can a complete audio picture be given. So maybe the 50% refers to first hour. If the complete CVR transcript takes less than two months, that will be rapid progress.

I find this quite intriguing. What is the source of this information?

I cannot remember the condition being given a name by the doctor I was sent to.
He asked me a series of questions one of which was 'Before puberty did you need to hold your nose when you swam underwater or jumped into water'
My answer was no.
He then asked me if I needed to do so now and I answered yes which was correct.
That was only part of the examination but he did not seem to think it was that unusual a condition.

I do not believe the condition can be picked up by normal hearing tests.

gcal: I think it's called "Sensorineural deafness".

January 28th date set to release preliminary report
 

Apparently, the radome of PK-AXC has been found washed up near Sembilan Island 550 KM away from the rest of the wreckage.

http://images.detik.com/content/2015...26_hidungc.jpg

Source: Detik.com

Approximate location of where the radome was found.

The name given is low levels of discrimination. The ability to make out a particular sound or voice in amongst other sounds, or even just the ability to understand speech.

In the extreme people can hear a voice but not actually discriminate the sounds into meaningful words. This is common in the deaf/hard of hearing and audiologists will carry out speech discrimination tests. This is usually done by playing voices of different people at different sound levels with or without background noise. They may talk in snippets or single words so there is no contextual clue to what the word is, or give a sentence that provides some contextual clues to what is being said.

The count the bleeps type hearing test is just for that hearing it does not identify people who cannot understand voices or sounds well against background noise.

I am surprised that this is not tested for. The person suffering a loss of speech or sound discrimination may not be aware of it, but will find talking/listening in noisy surroundings difficult and tiring.

This is totally different to the 'attentional' or 'cognitive' tunneling effect, where under pressure humans will focus on one particular part of their environment and exclude all others. If that is the wrong thing then that will lead to problems in a cockpit environment.
Sounds are one of the first stimuli to be filtered out, the last to be filtered out are haptics (touch and feeling) hence the reason for stick shakers.

@ Ian W - on transcripting
 

A0283 notes -- was my impression, and confirms other people talking about their experience with this hard work,

the captain involved taught one of the present investigators to fly, this investigator fully confirms your statements,

like i said in my post ... 'no professional would hold that 50% against them' ... And i was indeed impressed by the 50%

fully agree with that,

Note:

Once i took a number of tapes, listened to them, made a transcript, and compared that with what was published in appendices of the final reports. The easy part was that the report left out the private and human side ... In spite of that, you could feel the strain. The hard part was making a good transcript, my impression based on that is, that i would never leave it to a single person to make a transcription. And for more reasons than one, including sharing the tremendous emotional stress.

In the past you could find pretty complete tapes and texts. Today it appears that less and less is published. I wonder if that is wise. If you really want to understand an accident, then ... The negative side is of course people using the information with an uninformed or wrong intent.

There is no hail damage on that radome. It's in remarkably good condition. In fact it does not even show the type of erosion you normally see when a aircraft is flown through extreme rain.

So far there have been a lot of good suggestions re training for high altitude turbulence, stalling, and even spin recovery. However, as many have pointed out, the first priority is to avoid the extreme weather. There is a long history of jet fighters being lost in large CBs because they lacked any weather radar and yet they were stressed well above airliners' G limits and had much higher control response.

We can't properly simulate really severe turbulence, so the training - as for ditching - has to be rather arbitrary. There is a lot more we can do in regard to weather avoidance at reasonable cost/risk benefit by integrating weather data sources and I hope at least that this tragedy will spur greater efforts in that direction.

It has also become obvious that current automatics in extreme turbulence and icing may not cope adequately and that pilots are easily overstressed by violent movement and temporary or spurious warnings. Hopefully, this accident will provoke better algorithms in preference to relying more on manual takeovers. It's very hard in severe turbulence, blinded by lightning, to take over efficiently with your head and arms flailing around.

Having said that, each pilot must still know the limitations of their auto systems and be prepared for manual takeover. I'll bet on the technology in the long run. Self driving cars have, to my knowledge, not had an accident yet and had completed over 3 million miles last time I checked.

@ Ian W On sound and sound processing - under stress
 

Great post Ian W. It would not suprise me, based on the status by today, if these aspects will return in discussions following the investigation, and will even find their way into the final report.

When you go from relaxed to stressful situations, and end up at the most extreme stress levels on human beings - then you go through something that I call "roll back" (using my own words - I am sorry, but I do not know if there is a scientific label for the whole 'movement'). Which means that the brain's main processing moves from the frontal lobe, to the 'centre of the brain', and finally down to the 'brain stem'. One of the effects is for instance that you cannot lift as much weight in an extremely stressful environment as in a relaxed one. The reduction can be up to 50%. I do not know any numbers for the degradation of pull strength ( ref old fashioned sticks).

Under severe stress according to my psychiatric contacts you 'regress' towards primitive fight or flight (escape) mode. Both are problematic in high technology situations, which are not quite our evolutionary scenarios. One sub mode of 'flight' is to go catatonic or 'play possum'. I know of one incident where an inexperienced F/O got overwhelmed going into LHR and went to sleep and could not be woken until after landing.

On the topic of stress, there is an interesting Dutch investigation into military personnel. I believe though that the report might be of interest for anyone working in a high stress environment.

http://tinyurl.com/m9kwpga

Short summary of the above PDF: The "scientifical" words for what others call easy things like "roll-back" (I like that expression) is perceptual narrowing and indeed it messes up your perception. One of the first senses to be disregarded is hearing unless it is short, distinctive and preferrably in your mother tongue. Motor skills deteriorate because smaller muscles aren't provided with enough oxygen due to the fight-or-flight state that adrenaline dictates where it is most useful to have a good flow in your large muscles to be able to do either. Loss of near vision is also among the things that hampers your performance.

EDIT: Another one, more directly related to stress in aviation:

http://tinyurl.com/pflgb95

If the radome found belongs to the missing plane, it looks very clean and I see no visible hail marks or scorching as by lightning.

It doesn't look like it has hit the ocean at 100 knots either.

At that distance it's looking like it has detached before the crash, but whether it's turbulence or mechanical/latching failure remains to be determined.

Assuming the radome was attached to the aircraft when it hit the sea, the radome has drifted on average 24km per day (550/23).

Yes it could have drifted that distance and also no radome was reported lost when they requested climb only minutes before the crash position. It's possible also that it was detached in the uncontrolled descent if spinning, for example, with lateral airloads.

6000 fpm climb and subsequent stall reported BBC News - AirAsia flight QZ8501 'climbed too fast'

It's not a controlled rate of climb. 6000'/min is impossible at that altitude and type. It's a pitchup zoom climb possibly aided by updraft with reducing IAS into a stall; then we are guessing.